Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for enforcing Quality of Service (“QoS”) in a Virtual Service Network (“VSN”), comprising: defining, in a virtual layer, bandwidth values for each slice; weighting a first slice based on its bandwidth value relative to the bandwidth values of the other slices; receiving packets at a physical switch having ingress ports and egress ports; placing the packets into a slice-based pool having ingress queues, the ingress queues each corresponding to a slice in the VSN, wherein an agent running on the physical switch places the packets into the ingress queues based on slice classifications of the packets, wherein a first packet is classified differently than a second packet; routing the packets from the ingress queues to egress queues for the egress ports, the egress queues corresponding to the slices in the VSN; and forwarding the packets from the egress queues according to a policing algorithm, wherein the first packet is sent before the second packet based on having a prioritized slice classification, the prioritized slice classification being based on the weight of the first slice.
This invention relates to enforcing Quality of Service (QoS) in a Virtual Service Network (VSN) by managing bandwidth allocation and packet prioritization across network slices. The problem addressed is ensuring fair and efficient traffic handling in virtualized networks where multiple slices share physical infrastructure, often leading to congestion and inconsistent performance. The method involves defining bandwidth values for each network slice in a virtual layer, then weighting slices based on their allocated bandwidth relative to others. Packets are received at a physical switch with ingress and egress ports and classified into slice-based pools. Each slice has dedicated ingress queues, and an agent on the switch assigns packets to these queues based on their classifications. Packets from different slices may be classified differently, meaning they may be placed in different queues. The packets are then routed from ingress queues to egress queues corresponding to their slices. Finally, packets are forwarded from the egress queues using a policing algorithm that prioritizes packets from higher-weighted slices. For example, a packet from a prioritized slice will be sent before another packet from a lower-priority slice, ensuring that bandwidth allocation aligns with the defined weights. This approach ensures that network resources are allocated fairly and efficiently according to predefined QoS policies.
2. The method of claim 1 , wherein the first packet is classified into a first slice corresponding to emergency calls and the second packet is classified into a second slice corresponding to internet-of-things devices.
This invention relates to packet classification in wireless communication networks, specifically for prioritizing different types of traffic in network slicing environments. The problem addressed is the need to efficiently classify and route packets into distinct network slices to ensure quality of service (QoS) for critical services while optimizing resource allocation for less critical traffic. The method involves classifying packets into different network slices based on their type. For example, a first packet is classified into a first slice dedicated to emergency calls, ensuring low latency and high reliability for critical communications. A second packet is classified into a second slice designated for internet-of-things (IoT) devices, which may have different QoS requirements such as lower bandwidth but higher device density support. The classification ensures that emergency calls receive prioritized network resources, while IoT traffic is managed efficiently without degrading performance for higher-priority services. The method may also include additional steps such as monitoring network conditions, dynamically adjusting slice allocations, and enforcing QoS policies to maintain optimal performance across all slices. This approach enables network operators to support diverse services with varying requirements within a shared infrastructure.
3. The method of claim 1 , wherein subsequent packets in a same stream as the first packet are classified into the same slice as the first packet based on being part of the same stream.
This invention relates to network packet classification for efficient processing in a multi-slice system. The problem addressed is the need to classify packets into appropriate processing slices to optimize resource utilization and performance in network traffic management. The method involves classifying packets into slices based on their stream membership, ensuring packets from the same stream are processed in the same slice. A first packet is initially classified into a slice, and subsequent packets belonging to the same stream are automatically assigned to the same slice. This classification is based on stream identification, which may involve analyzing packet headers or other identifying features. The method improves efficiency by reducing inter-slice communication and ensuring consistent processing for related packets. The system may include multiple processing slices, each handling a subset of network traffic, with classification rules dynamically updated to adapt to changing network conditions. The invention is particularly useful in high-performance networking environments where low-latency and high-throughput processing are critical.
4. The method of claim 1 , wherein placing the packets from the ingress queues into egress queues is done based on a first-in-first-out policy.
This invention relates to packet processing in network systems, specifically addressing the challenge of efficiently managing packet flow between ingress and egress queues to optimize network performance. The method involves transferring packets from ingress queues to egress queues using a first-in-first-out (FIFO) policy. This ensures that packets are processed in the order they are received, preventing starvation and maintaining fairness in packet delivery. The system includes multiple ingress queues for receiving packets from different sources and multiple egress queues for forwarding packets to their destinations. The FIFO policy applies to the distribution of packets from the ingress queues to the egress queues, ensuring that the oldest packets are prioritized. This approach helps maintain low latency and consistent throughput, particularly in high-traffic scenarios where packet ordering is critical. The method may also include additional steps such as monitoring queue lengths and dynamically adjusting the distribution process to prevent congestion. By enforcing a strict FIFO policy, the invention ensures predictable and reliable packet delivery, which is essential for real-time applications and services. The system may be implemented in network switches, routers, or other packet-processing devices to enhance overall network efficiency.
5. The method of claim 4 , wherein the policing algorithm is a slice weighted round robin algorithm.
A method for managing network traffic in a wireless communication system, particularly in a 5G or similar network architecture, addresses the challenge of efficiently allocating resources among multiple network slices. Network slicing allows different services to share the same physical infrastructure while maintaining isolation and quality of service (QoS) guarantees. The method involves dynamically allocating resources to different network slices based on their priority and demand, ensuring that high-priority slices receive sufficient bandwidth while lower-priority slices are allocated remaining resources. The method employs a policing algorithm to enforce these allocations, preventing any single slice from monopolizing network resources. Specifically, the policing algorithm is a slice-weighted round robin algorithm, which distributes resources in a cyclic manner but adjusts the allocation weights based on predefined priorities or service-level agreements (SLAs). This ensures fairness and prevents starvation of lower-priority slices while maintaining performance for critical services. The algorithm operates by assigning weights to each slice, where higher-priority slices receive larger weights, and then cycling through the slices in a round-robin fashion, allocating resources proportionally to their weights. This approach balances efficiency and fairness, dynamically adapting to changing network conditions and traffic patterns. The method is particularly useful in scenarios where multiple slices with varying QoS requirements coexist, such as in mobile broadband, IoT, or mission-critical communications.
6. The method of claim 1 , further comprising: weighting a second slice based on its bandwidth value relative to the bandwidth values of the other slices; and based on the weight of the first slice being greater than the weight of the second slice, sending the first packet before the second packet.
This invention relates to packet scheduling in communication systems, particularly for prioritizing packet transmission based on bandwidth allocation. The problem addressed is efficient and fair allocation of network resources when multiple data slices (logical partitions of bandwidth) compete for transmission. The method involves dynamically weighting slices according to their allocated bandwidth values, ensuring higher-priority slices receive preferential treatment. When a first packet from a higher-weighted slice and a second packet from a lower-weighted slice are ready for transmission, the system sends the first packet before the second, even if the second packet was ready earlier. This ensures that bandwidth allocations are respected, preventing lower-priority slices from monopolizing transmission resources. The weighting mechanism may involve comparing the bandwidth values of all active slices to determine relative priorities. The method is particularly useful in systems where multiple users or services share a limited bandwidth resource, such as wireless networks or data centers, to maintain fairness and service-level agreements. The invention improves network efficiency by aligning transmission order with allocated bandwidth, reducing latency for high-priority traffic while ensuring lower-priority traffic still receives its fair share.
7. The method of claim 1 , further comprising: installing the agent at the physical switch, wherein the agent performs the slice classification based on packet header information including at least a source address, destination address, and destination port.
This invention relates to network traffic management in software-defined networking (SDN) environments, specifically for classifying and managing network slices based on packet header information. The problem addressed is the need for efficient and accurate classification of network traffic to support multiple virtualized network slices within a shared physical infrastructure. Traditional methods often rely on complex rule sets or centralized controllers, which can introduce latency and scalability issues. The invention involves a method for classifying network traffic into predefined slices using an agent installed at a physical switch. The agent performs slice classification by analyzing packet header information, including the source address, destination address, and destination port. This allows the switch to dynamically assign traffic to the appropriate network slice without requiring constant intervention from a central controller. The classification process is performed locally at the switch, reducing latency and improving scalability. The agent can also enforce policies or rules associated with each slice, ensuring proper traffic isolation and quality of service (QoS) guarantees. This approach enables efficient multi-tenancy and service differentiation in SDN environments while minimizing the overhead on the central controller. The invention improves network performance by decentralizing slice classification and leveraging local processing capabilities at the switch level.
8. A non-transitory, computer-readable medium comprising instructions that, when executed by a processor, perform stages for enforcing Quality of Service (“QoS”) in a Virtual Service Network (“VSN”), the stages comprising: defining, in a virtual layer, bandwidth values for each slice; weighting a first slice based on its bandwidth value relative to the bandwidth values of the other slices; receiving packets at a physical switch having ingress ports and egress ports; placing the packets into a slice-based pool having ingress queues, the ingress queues each corresponding to a slice in the VSN, wherein an agent running on the physical switch places the packets into the ingress queues based on slice classifications of the packets, wherein a first packet is classified differently than a second packet; routing the packets from the ingress queues to egress queues for the egress ports, the egress queues corresponding to the slices in the VSN; and forwarding the packets from the egress queues according to a policing algorithm, wherein the first packet is sent before the second packet based on having a prioritized slice classification, the prioritized slice classification being based on the weight of the first slice.
This invention relates to enforcing Quality of Service (QoS) in a Virtual Service Network (VSN) by managing bandwidth allocation and packet prioritization across network slices. The system operates in a virtual layer to define bandwidth values for each network slice, then weights slices based on their relative bandwidth allocations. A physical switch with ingress and egress ports receives packets and classifies them into slice-based ingress queues, where packets are sorted by their slice classifications. An agent running on the switch ensures proper classification. Packets are then routed from ingress queues to egress queues corresponding to their respective slices. Finally, a policing algorithm forwards packets from the egress queues, prioritizing packets from higher-weighted slices. For example, a first packet classified in a prioritized slice is forwarded before a second packet in a lower-priority slice, based on the weighted bandwidth values. This approach ensures that network resources are allocated according to predefined QoS policies, improving traffic management in virtualized networks.
9. The non-transitory, computer-readable medium of claim 8 , wherein the first packet is classified into a first slice corresponding to emergency calls and the second packet is classified into a second slice corresponding to internet-of-things devices.
This invention relates to network traffic classification in a telecommunications system, specifically for prioritizing different types of data packets based on their intended use. The system classifies packets into distinct network slices to ensure efficient resource allocation and prioritization. The invention addresses the challenge of managing diverse traffic types in modern networks, where different applications require varying levels of latency, reliability, and bandwidth. The system processes data packets by analyzing their characteristics to determine their intended use. For example, a first packet is classified into a network slice dedicated to emergency calls, ensuring low-latency and high-priority handling. A second packet is classified into a separate slice for internet-of-things (IoT) devices, which may require different quality-of-service parameters. The classification ensures that emergency communications receive immediate attention while IoT traffic is managed according to its specific needs, such as periodic data transmission or low-power operation. The invention improves network efficiency by dynamically allocating resources based on packet classification, reducing congestion and optimizing performance for different traffic types. This approach is particularly useful in 5G and beyond networks, where multiple services with varying requirements coexist. The system may also include additional classification rules for other packet types, ensuring comprehensive traffic management.
10. The non-transitory, computer-readable medium of claim 8 , wherein subsequent packets in a same stream as the first packet are classified into the same slice as the first packet based on being part of the same stream.
A system and method for classifying network packets into processing slices based on stream membership. The technology addresses the challenge of efficiently distributing network traffic across multiple processing units in a scalable manner. Traditional packet classification methods often rely on complex rule sets or deep packet inspection, which can introduce latency and processing overhead. This invention improves upon prior approaches by dynamically assigning packets to processing slices based on their stream membership, ensuring that all packets belonging to the same communication stream are processed by the same slice. This reduces synchronization overhead between processing units and improves throughput. The system first identifies a stream associated with a packet, such as a TCP or UDP flow, and then classifies subsequent packets in that stream into the same processing slice. The classification is based on stream identifiers, such as source and destination IP addresses, port numbers, and protocol types. This method ensures consistent processing of related packets while minimizing inter-slice communication. The invention is particularly useful in high-performance networking environments, such as data centers, cloud computing, and network security appliances, where efficient packet processing is critical. The solution enhances scalability and reduces processing latency by leveraging stream-based classification rather than per-packet analysis.
11. The non-transitory, computer-readable medium of claim 8 , wherein placing the packets from the ingress queues into egress queues is done based on a first-in-first-out policy.
A system and method for managing packet processing in a network device involves distributing packets from ingress queues to egress queues using a first-in-first-out (FIFO) policy. The system includes a network device with multiple ingress and egress queues, where packets are received at the ingress queues and then distributed to the appropriate egress queues for transmission. The distribution process ensures that packets are processed in the order they were received, maintaining fairness and preventing starvation of any particular packet flow. The system may also include mechanisms for monitoring queue lengths and dynamically adjusting the distribution policy to optimize performance and minimize latency. This approach is particularly useful in high-speed networking environments where maintaining strict ordering of packets is critical for reliable communication. The FIFO policy ensures that packets are processed sequentially, reducing the risk of packet reordering, which can degrade network performance. The system may further include error detection and correction mechanisms to handle packet loss or corruption during transmission. By implementing a FIFO-based distribution policy, the system provides a robust and efficient method for managing packet flow in network devices, ensuring consistent and predictable performance.
12. The non-transitory, computer-readable medium of claim 11 , wherein the policing algorithm is a slice weighted round robin algorithm.
A system and method for managing network traffic in a wireless communication environment, particularly in scenarios involving network slicing, where multiple virtual networks share physical infrastructure. The problem addressed is ensuring fair and efficient allocation of network resources among different network slices, each with varying quality of service (QoS) requirements. The solution involves a policing algorithm implemented on a non-transitory, computer-readable medium to regulate traffic flow between a user equipment (UE) and a network. The algorithm dynamically adjusts resource allocation based on predefined policies, ensuring that each slice receives its allocated share of bandwidth while preventing any single slice from monopolizing resources. The policing algorithm operates by monitoring traffic patterns, applying predefined rules, and enforcing compliance with service level agreements (SLAs). In one embodiment, the policing algorithm is a slice-weighted round robin algorithm, which prioritizes traffic distribution in a cyclical manner, weighted by the importance or priority of each slice. This ensures that higher-priority slices receive more frequent access to network resources while still allowing lower-priority slices to maintain minimum service levels. The system may also include mechanisms for real-time adjustments based on network conditions, such as congestion or latency, to optimize overall performance. The solution is particularly useful in 5G and beyond networks, where network slicing is a key feature for supporting diverse applications with varying QoS demands.
13. The non-transitory, computer-readable medium of claim 8 , the stages further comprising: weighting a second slice based on its bandwidth value relative to the bandwidth values of the other slices; and based on the weight of the first slice being greater than the weight of the second slice, sending the first packet before the second packet.
This invention relates to packet scheduling in network communication systems, specifically addressing the challenge of efficiently prioritizing packet transmission based on bandwidth allocation. The system dynamically assigns weights to data slices (segments of network traffic) according to their allocated bandwidth values, ensuring higher-priority transmission for slices with greater bandwidth. When two packets from different slices are ready for transmission, the system compares their weights. If the first slice's weight exceeds the second slice's weight, the first packet is sent before the second, optimizing bandwidth utilization and reducing latency for higher-priority traffic. The method involves monitoring bandwidth allocations, calculating relative weights, and enforcing transmission order based on these weights. This approach improves fairness and efficiency in network resource allocation, particularly in environments with variable bandwidth demands. The invention is implemented via a computer-readable medium containing instructions for executing these stages, ensuring adaptability to different network conditions. The solution enhances performance by dynamically adjusting transmission priorities without requiring static configurations, making it suitable for modern, dynamic network architectures.
14. The non-transitory, computer-readable medium of claim 8 , the stages further comprising: installing the agent at the physical switch, wherein the agent performs the slice classification based on packet header information including at least a source address, destination address, and destination port.
This invention relates to network traffic management in software-defined networking (SDN) environments, specifically for classifying and managing network traffic slices based on packet header information. The problem addressed is the need for efficient and scalable traffic classification in high-performance network switches to support multi-tenant or service-specific network slicing. The invention involves a non-transitory computer-readable medium storing instructions for a network controller to configure a physical switch with an agent. The agent is installed on the physical switch and performs slice classification by analyzing packet header information, including the source address, destination address, and destination port. This classification enables the switch to dynamically assign traffic to different network slices, ensuring proper isolation, prioritization, and quality of service (QoS) for different types of traffic. The agent operates at the switch level, reducing the need for centralized processing and improving classification speed and efficiency. By leveraging packet header fields, the system can accurately categorize traffic without deep packet inspection, minimizing latency and computational overhead. This approach is particularly useful in environments where multiple virtual networks or services share the same physical infrastructure, such as data centers, cloud computing, or 5G networks. The solution enhances scalability and flexibility in network management while maintaining performance and security.
15. A system for enforcing Quality of Service (“QoS”) in a Virtual Service Network (“VSN”), comprising: a non-transitory, computer-readable medium containing instructions; and a processor that executes the instructions to perform stages comprising: defining, in a virtual layer, bandwidth values for each slice; weighting a first slice based on its bandwidth value relative to the bandwidth values of the other slices; receiving packets at a physical switch having ingress ports and egress ports; placing the packets into a slice-based pool having ingress queues, the ingress queues each corresponding to a slice in the VSN, wherein an agent running on the physical switch places the packets into the ingress queues based on slice classifications of the packets, wherein a first packet is classified differently than a second packet; routing the packets from the ingress queues to egress queues for the egress ports, the egress queues corresponding to the slices in the VSN; and forwarding the packets from the egress queues according to a policing algorithm, wherein the first packet is sent before the second packet based on having a prioritized slice classification, the prioritized slice classification being based on the weight of the first slice.
The system enforces Quality of Service (QoS) in a Virtual Service Network (VSN) by managing bandwidth allocation and packet prioritization across multiple network slices. The system operates in a virtual layer, where bandwidth values are defined for each slice, and slices are weighted based on their relative bandwidth allocations. A physical switch with ingress and egress ports receives packets and classifies them into slice-based ingress queues, with each queue corresponding to a specific slice. An agent on the physical switch ensures packets are placed into the correct queues based on their classifications, allowing different packets to be assigned to different slices. Packets are then routed from ingress queues to egress queues, which are also slice-specific. The system forwards packets from the egress queues using a policing algorithm that prioritizes packets based on their slice classifications. A packet from a higher-weighted slice is forwarded before a packet from a lower-weighted slice, ensuring that bandwidth is allocated according to the predefined slice weights. This approach enables efficient QoS enforcement by dynamically prioritizing traffic based on slice importance and bandwidth allocation.
16. The system of claim 15 , wherein the first packet is classified into a first slice corresponding to emergency calls and the second packet is classified into a second slice corresponding to internet-of-things devices.
This invention relates to a network system that classifies and processes data packets based on their type or priority. The system is designed to handle different types of network traffic efficiently, particularly in scenarios where multiple traffic types must be managed simultaneously. The system includes a classification module that categorizes incoming packets into distinct network slices, each optimized for specific use cases. One slice is dedicated to emergency calls, ensuring high-priority, low-latency processing to support critical communications. Another slice is allocated for internet-of-things (IoT) devices, which may require different handling due to their unique traffic patterns, such as periodic data transmissions or low-bandwidth requirements. The system dynamically assigns resources to each slice based on demand, ensuring that emergency traffic receives priority while IoT traffic is managed efficiently without unnecessary resource allocation. This approach improves network reliability and performance by isolating different traffic types, preventing congestion, and optimizing resource usage. The system may also include additional slices for other traffic types, such as general internet access or voice communications, further enhancing flexibility. The classification module uses predefined rules or machine learning to determine the appropriate slice for each packet, ensuring accurate and rapid processing.
17. The system of claim 15 , wherein subsequent packets in a same stream as the first packet are classified into the same slice as the first packet based on being part of the same stream.
This invention relates to network traffic classification and processing, specifically for efficiently managing data streams in a network system. The problem addressed is the need to classify and process packets belonging to the same data stream consistently, ensuring that packets in a stream are handled uniformly to maintain performance and security. The system includes a network traffic classifier that processes incoming packets, including a first packet in a data stream. The classifier determines a classification for the first packet, such as identifying its type, source, or priority. The system then assigns the first packet to a specific processing slice, which is a logical or physical partition for handling packets with similar characteristics. Subsequent packets in the same stream as the first packet are automatically classified into the same slice as the first packet, based on their association with the same stream. This ensures that all packets in a stream are processed in a consistent manner, improving efficiency and reducing misclassification errors. The system may also include mechanisms to detect and handle stream changes, such as when a stream's characteristics or requirements alter, requiring reclassification. The invention enhances network performance by reducing processing overhead and ensuring predictable handling of related packets.
18. The system of claim 15 , wherein placing the packets from the ingress queues into egress queues is done based on a first-in-first-out policy.
A system for managing packet processing in a network device addresses the challenge of efficiently distributing packets from ingress queues to egress queues to optimize throughput and reduce latency. The system includes a packet processor that receives packets from multiple ingress queues and distributes them to multiple egress queues based on a scheduling policy. The scheduling policy determines the order in which packets are selected from the ingress queues and placed into the egress queues. In this specific implementation, the scheduling policy enforces a first-in-first-out (FIFO) policy, ensuring that packets are processed in the order they arrive. This approach minimizes reordering delays and maintains fairness in packet transmission. The system may also include mechanisms to monitor queue depths and dynamically adjust the scheduling policy to prevent congestion. By using a FIFO policy, the system ensures predictable and consistent packet delivery, which is critical for real-time applications and quality-of-service guarantees. The system is particularly useful in high-speed networking environments where low latency and high throughput are required.
19. The system of claim 18 , wherein the policing algorithm is a slice weighted round robin algorithm.
A system for managing network traffic in a wireless communication environment, particularly in scenarios involving network slicing, where multiple virtual networks share the same physical infrastructure. The system addresses the challenge of ensuring fair and efficient resource allocation among different network slices, each with varying quality of service (QoS) requirements. The system includes a policing algorithm that dynamically allocates bandwidth and other resources to different slices based on predefined policies and real-time network conditions. This ensures that critical slices receive priority while maintaining overall network stability. The policing algorithm is specifically implemented as a slice-weighted round robin algorithm, which distributes resources in a cyclical manner but adjusts the allocation weights according to the priority and requirements of each slice. This approach prevents any single slice from monopolizing resources while optimizing overall network performance. The system may also include monitoring components to track slice performance and adjust the policing algorithm in real-time to adapt to changing demands. The solution is particularly useful in 5G and beyond networks, where network slicing is a key feature for supporting diverse services with different QoS needs.
20. The system of claim 15 , the stages further comprising: weighting a second slice based on its bandwidth value relative to the bandwidth values of the other slices; and based on the weight of the first slice being greater than the weight of the second slice, sending the first packet before the second packet.
This invention relates to a system for prioritizing packet transmission in a network, particularly in scenarios where multiple data slices compete for bandwidth. The problem addressed is inefficient bandwidth utilization and packet delivery delays when multiple data slices with varying bandwidth requirements are processed without proper prioritization. The system dynamically assigns weights to data slices based on their bandwidth values relative to other slices, ensuring higher-priority slices receive preferential treatment. When a first packet from a higher-weighted slice is ready for transmission alongside a second packet from a lower-weighted slice, the system sends the first packet before the second packet, optimizing bandwidth allocation and reducing latency for critical data. The system may also include mechanisms for determining slice bandwidth values, managing packet queues, and dynamically adjusting weights based on real-time network conditions. This approach improves overall network efficiency by ensuring that higher-bandwidth slices are serviced first, minimizing delays for time-sensitive applications. The invention is particularly useful in networks with heterogeneous traffic demands, such as those handling both high-priority real-time data and lower-priority background traffic.
Unknown
January 12, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.